With this latest partnership, combined with the Stargate I site in Abilene, Texas, OpenAI now reports that it has over 5 gigawatts of Stargate AI data center capacity under development. These data centers are expected to operate more than two million compute chips, which OpenAI says will significantly increase its ability to run large-scale AI training and inference operations.
OpenAI claims that these initiatives align with its previously announced commitment to invest $500 billion in 10 gigawatts of AI infrastructure in the US over four years—a goal it now expects to exceed, citing strong support from partners like Oracle and SoftBank. The company estimates the construction, development, and operation of the additional 4.5 gigawatts of capacity will create over 100,000 jobs in US construction, operations, manufacturing, and related local services.
OpenAI reports that construction of the Stargate I data center is ongoing, with portions of the facility operational and initial Nvidia GB200 server racks delivered by Oracle. OpenAI states that these new resources are being used for early-stage training and inference tasks, supporting research into next-generation AI models. The Stargate I site, according to the company, has already generated thousands of specialized jobs, including roles for electricians, operators, and technicians.
OpenAI also notes active progress on its partnership with SoftBank, aimed at accelerating site assessments and data center designs optimized for advanced AI workloads. The overall Stargate platform encompasses collaborations with Oracle, SoftBank, and CoreWeave, as well as international initiatives through OpenAI for Countries. Microsoft will continue supplying cloud services to OpenAI, including through Stargate.
Source: OpenAI
